The role of water in protein-DNA recognition

Entropy of Water in the Hydration Layer of Major and Minor Grooves of DNA

Transport properties (translational and rotational) of water in the two grooves of the B-DNA duplex are known to be different from those in the bulk. Here, we use a recently developed theoretical scheme to compute the entropies of water molecules in both of the grooves of DNA and compare them with that in the bulk. The scheme requires as input both translational and rotational velocity autocorrelation function (C(V)(t) and C(omega)(t), respectively) data. These velocity autocorrelation functions were computed from an atomistic MD simulation of a B-DNA duplex (36 base pairs long) in explicit water (TIP3P). The average values of the entropy of water at 300 K in both of the grooves of DNA (the TS value in the major groove is 6.71 kcal/mol and that in the minor groove is 6.41 kcal/mol) are found to be significantly lower than that in bulk water (the TS value is 7.27 kcal/mol). Thus, the entropic contribution to the free energy change (TDeltaS) of transferring a minor groove water molecule to the bulk is 0.86 kcal/mol and of transferring a major groove water to the bulk is 0.56 kcal/mol at 300 K, which is to be compared with 1.44 kcal/mol for melting of ice at 273 K. We also calculate the energy of interaction of each water molecule with the rest of the atoms in the system and hence calculate the chemical potential (Helmholtz free energy per water molecule, A = E - TS) in the different domains. The identical free energy value of water molecules in the different domains proves the robustness of the scheme. We propose that the configurational entropy of water in the grooves can be used as a measure of the mobility (or microviscosity) of water molecules in a given domain.

Structural basis for polypyrimidine tract recognition by the essential pre-mRNA splicing factor U2AF65

The essential pre-mRNA splicing factor, U2AF(65), guides the early stages of splice site choice by recognizing a polypyrimidine (Py) tract consensus sequence near the 3' splice site. Since Py tracts are relatively poorly conserved in higher eukaryotes, U2AF(65) is faced with the problem of specifying uridine-rich sequences, yet tolerating a variety of nucleotide substitutions found in natural Py tracts. To better understand these apparently contradictory RNA binding characteristics, the X-ray structure of the U2AF(65) RNA binding domain bound to a Py tract composed of seven uridines has been determined at 2.5 A resolution. Specific hydrogen bonds between U2AF(65) and the uracil bases provide an explanation for polyuridine recognition. Flexible side chains and bound water molecules form the majority of the base contacts and potentially could rearrange when the U2AF(65) structure adapts to different Py tract sequences. The energetic importance of conserved residues for Py tract binding is established by analysis of site-directed mutant U2AF(65) proteins using surface plasmon resonance.

Sequestered water and binding energy are coupled in complexes of lambda Cro repressor with non-consensus binding sequences

We use the osmotic pressure dependence of dissociation rates and relative binding constants to infer differences in sequestered water among complexes of lambda Cro repressor with varied DNA recognition sequences. For over a 1000-fold change in association constant, the number of water molecules sequestered by non-cognate complexes varies linearly with binding free energy. One extra bound water molecule is coupled with the loss of approximately 150 cal/mol complex in binding free energy. Equivalently, every tenfold decrease in binding constant at constant salt and temperature is associated with eight to nine additional water molecules sequestered in the non-cognate complex. The relative insensitivity of the difference in water molecules to the nature of the osmolyte used to probe the reaction suggests that the water is sterically sequestered. If the previously measured changes in heat capacity for lambda Cro binding to different non-cognate sequences are attributed solely to this change in water, then the heat capacity change per incorporated water is almost the same as the difference between ice and water. The associated changes in enthalpies and entropies, however, indicate that the change in complex structure involves more than a simple incorporation of fixed water molecules that act as adaptors between non-complementary surfaces.

Binding of netropsin to several DNA constructs: evidence for at least two different 1:1 complexes formed from an -AATT-containing ds-DNA construct and a single minor groove binding ligand

Isothermal titration calorimetry, ITC, has been used to determine the thermodynamics (DeltaG, DeltaH, and -TDeltaS) for binding netropsin to a number of DNA constructs. The DNA constructs included: six different 20-22mer hairpin forming sequences and an 8-mer DNA forming a duplex dimer. All DNA constructs had a single -AT-rich netropsin binding with one of the following sequences, (A(2)T(2))(2), (ATAT)(2), or (AAAA/TTTT). Binding energetics are less dependent on site sequence than on changes in the neighboring single stranded DNA (hairpin loop size and tail length). All of the 1:1 complexes exhibit an enthalpy change that is dependent on the fractional saturation of the binding site. Later binding ligands interact with a significantly more favorable enthalpy change (partial differential DeltaH(1-2) from 2 to 6 kcal/mol) and a significantly less favorable entropy change (partial differential (-TDeltaS(1-2))) from -4 to -9 kcal/mol). The ITC data could only be fit within expected experimental error by use of a thermodynamic model that includes two independent binding processes with a combined stoichiometry of 1 mol of ligand per 1 mol of oligonucleotide. Based on the biophysical evidence reported here, including theoretical calculations for the energetics of "trapping" or structuring of a single water molecule and molecular docking computations, it is proposed that there are two modes by which flexible ligands can bind in the minor groove of duplex DNA. The higher affinity binding mode is for netropsin to lay along the floor of the minor groove in a bent conformation and exclude all water from the groove. The slightly weaker binding mode is for the netropsin molecule to have a slightly more linear conformation and for the required curvature to be the result of a water molecule that bridges between the floor of the minor groove and two of the amidino nitrogens located at one end of the bound netropsin molecule.

WATER MEDIATION IN PROTEIN FOLDING AND MOLECULAR RECOGNITION

Water is essential for life in many ways, and without it biomolecules might no longer truly be biomolecules. In particular, water is important to the structure, stability, dynamics, and function of biological macromolecules. In protein folding, water mediates the collapse of the chain and the search for the native topology through a funneled energy landscape. Water actively participates in molecular recognition by mediating the interactions between binding partners and contributes to either enthalpic or entropic stabilization. Accordingly, water must be included in recognition and structure prediction codes to capture specificity. Thus water should not be treated as an inert environment, but rather as an integral and active component of biomolecular systems, where it has both dynamic and structural roles. Focusing on water sheds light on the physics and function of biological machinery and self-assembly and may advance our understanding of the natural design of proteins and nucleic acids.

Weight matrices for protein-DNA binding sites from a single co-crystal structure

Transcription-factor proteins bind to specific DNA sequences to regulate gene expression in cells. DNA-binding sites are often identified using weight matrices calculated from multiple known binding sites. However, in many cases the number of examples is limited. Here, we report on an atomistic method that starts from an x-ray co-crystal structure of the protein bound to one particular DNA sequence, and infers other binding sites, which are used to construct a weight matrix. The emphasis of the paper is on using the Wang-Landau Monte Carlo algorithm to efficiently sample high-affinity binding sites, which demonstrates that sampling can produce accurate weight matrices in analogy to bioinformatics approaches. For cases of low complexity, we compare to the exhaustive (but slow) dead-end elimination algorithm. To recover crystal binding sites, it is important to include bound water in the protein-DNA interface. Our approach can, in principle, even be applied when no native protein-DNA co-crystal structure is available, only the structure of a closely related homologous protein whose amino-acid sequence is changed to the protein of interest.

Directed evolution and substrate specificity profile of homing endonuclease I-SceI

The laboratory evolution of enzymes with tailor-made DNA cleavage specificities would represent new tools for manipulating genomes and may enhance our understanding of sequence-specific DNA recognition by nucleases. Below we describe the development and successful application of an efficient in vivo positive and negative selection system that applies evolutionary pressure either to favor the cleavage of a desired target sequence or to disfavor the cleavage of nontarget sequences. We also applied a previously described in vitro selection method to reveal the comprehensive substrate specificity profile of the wild-type I-SceI homing endonuclease. Together these tools were used to successfully evolve mutant I-SceI homing endonucleases with altered DNA cleavage specificities. The most highly evolved enzyme cleaves the target mutant DNA sequence with a selectivity that is comparable to wild-type I-SceI's preference for its cognate substrate.

Protein-nucleic acid recognition: statistical analysis of atomic interactions and influence of DNA structure

We analyzed structural features of 11,038 direct atomic contacts (either electrostatic, H-bonds, hydrophobic, or other van der Waals interactions) extracted from 139 protein-DNA and 49 protein-RNA nonhomologous complexes from the Protein Data Bank (PDB). Globally, H-bonds are the most frequent interactions (approximately 50%), followed by van der Waals, hydrophobic, and electrostatic interactions. From the protein viewpoint, hydrophilic amino acids are over-represented in the interaction databases: Positively charged amino acids mainly contact nucleic acid phosphate groups but can also interact with base edges. From the nucleotide point of view, DNA and RNA behave differently: Most protein-DNA interactions involve phosphate atoms, while protein-RNA interactions involve more frequently base edge and ribose atoms. The increased participation of DNA phosphate involves H-bonds rather than salt bridges. A statistical analysis was performed to find the occurrence of amino acid-nucleotide pairs most different from chance. These pairs were analyzed individually. Finally, we studied the conformation of DNA in the interaction sites. Despite the prevalence of B-DNA in the database, our results suggest that A-DNA is favored in the interaction sites.

Solvent participation in Serratia marcescens endonuclease complexes

The monomer and dimer of the bacterium Serratia marcescens endonuclease (SMnase) are each catalytically active and the two subunits of the dimer function independently of each other. Specific interfacial waters may play a role in stability, complex formation, and functionality. We performed molecular dynamics simulations of both a subunit of SMnase and its model built complex with DNA and analyzed the relation of the hydration sites to the catalytic mechanism. It was found that the binding of DNA has little influence on the global hydration properties of the protein, including occupancy and water residence time distributions. DNA and protein recognition in our model mainly involves direct contacts by hydrogen bond and hydrophobic interactions. Water-mediated contacts exist, but are less common. Three interior water clusters were identified for SMnase. One cluster around the active site in the monomer-DNA complex shows relatively strong interactions between hydration sites as well as between the sites and the biomolecules. The simulated cluster properties agreed well with experimental data. The magnesium ion shows ligand exchange. Although Mg2+ keeps six ligands during the entire simulation, upon the binding of DNA, Asn119 loses its coordination with Mg2+, while one nonbridging oxygen of the phosphate of a DNA residue and two oxygen atoms of the side chain of Glu127 become the ligands. Waters in a nearby cluster exchange and participate in the resolvation of groups in the presence of DNA. Water thus not only participates in the cleavage of DNA but also can stabilize the transition state and the leaving groups in our model.

Water structure and interactions with protein surfaces

The structure of liquid water and its interaction with biological molecules is a very active area of experimental and theoretical research. The chemically complex surfaces of protein molecules alter the structure of the surrounding layer of hydrating water molecules. The dynamics of hydration water can be detected by a series of experimental techniques, which show that hydration waters typically have slower correlation times than water in bulk. Specific water-mediated interactions in protein complexes have been studied in detail, and these interactions have been incorporated into potential energy functions for protein folding and design. The subtle changes in the structure of hydration water have been investigated by theoretical studies.

Efficient prediction of nucleic acid binding function from low-resolution protein structures

Structural genomics projects as well as ab initio protein structure prediction methods provide structures of proteins with no sequence or fold similarity to proteins with known functions. These are often low-resolution structures that may only include the positions of C alpha atoms. We present a fast and efficient method to predict DNA-binding proteins from just the amino acid sequences and low-resolution, C alpha-only protein models. The method uses the relative proportions of certain amino acids in the protein sequence, the asymmetry of the spatial distribution of certain other amino acids as well as the dipole moment of the molecule. These quantities are used in a linear formula, with coefficients derived from logistic regression performed on a training set, and DNA-binding is predicted based on whether the result is above a certain threshold. We show that the method is insensitive to errors in the atomic coordinates and provides correct predictions even on inaccurate protein models. We demonstrate that the method is capable of predicting proteins with novel binding site motifs and structures solved in an unbound state. The accuracy of our method is close to another, published method that uses all-atom structures, time-consuming calculations and information on conserved residues.

High-resolution protein hydration NMR experiments: Probing how protein surfaces interact with water and other non-covalent ligands

High-resolution solution NMR experiments are extremely useful to characterize the location and the dynamics of hydrating water molecules at atomic resolution. However, these methods are severely limited by undesired incoherent transfer pathways such as those arising from exchange-relayed intra-molecular cross-relaxation. Here, we review several complementary exchange network editing methods that can be used in conjunction with other types of NMR hydration experiments such as magnetic relaxation dispersion and 1J(NC') measurements to circumvent these limitations. We also review several recent contributions illustrating how the original solution hydration NMR pulse sequence architecture has inspired new approaches to map other types of non-covalent interactions going well beyond the initial scope of hydration. Specifically, we will show how hydration NMR methods have evolved and have been adapted to binding site mapping, ligand screening, protein-peptide and peptide-lipid interaction profiling.

Hydration and conformational transitions in DNA, RNA, and mixed DNA-RNA triplexes studied by gravimetry and FTIR spectroscopy

We have studied by gravimetric measurements and FTIR spectroscopy the hydration of duplexes and triplexes formed by combinations of dA(n), dT(n), rA(n), and rU(n) strands. Results obtained on hydrated films show important differences in their hydration and in the structural transitions which can be induced by varying the water content of the samples. The number of water molecules per nucleotide (w/n) measured at high relative humidity (98% R.H.) is found to be 21 for dA(n).dT(n) and 15 for rA(n).rU(n). Addition of a third rU(n) strand does not change the number of water molecules per nucleotide: w/n=21 for rU(n)*dA(n).dT(n) and w/n=15 for rU(n)*rA(n).rU(n). On the contrary, the addition of a third dT(n) strand changes the water content but in a different way, depending whether the duplex is DNA or RNA. Thus, a loss of four water molecules per nucleotide is measured for dT(n)*dA(n).dT(n) while an increase of two water molecules per nucleotide is observed for dT(n)*rA(n).rU(n). The final hydration is the same for both triplexes (w/n=17). The desorption profiles obtained by gravimetry and FTIR spectroscopy are similar for the rA(n).rU(n) duplex and the rU(n)*rA(n).rU(n) triplex. On the contrary, the desorption profiles of the dA(n).dT(n) duplex and the triplexes formed with it (rU(n)*dA(n).dT(n) and dT(n)*dA(n).dT(n)) are different from each other. This is correlated with conformational transitions induced by varying the hydration content of the different structures, as shown by FTIR spectroscopy. Modifications of the phosphate group hydration and of the sugar conformation (S to N type repuckering) induced by decrease of the water content are observed in the case of triplexes formed on the dA(n).dT(n) duplex.

Hydration changes in the association of Hoechst 33258 with DNA

The role of water in the interaction of Hoechst 33258 with the minor groove binding site of the (AATT)2 sequence was investigated using calorimetric and equilibrium constant measurements. Using isothermal titration calorimetry measurements, the heat capacity change for the reaction is -256 +/- 10 cal/(K mol of Hoechst). Comparison with the heat capacity changes based on area models supports the expulsion of water from the interface of the Hoechst-DNA complex. To further consider the role of water, the osmotic stress method was used to determine if the Hoechst association with DNA was coupled with hydration changes. Using four osmolytes with varying molecular weights and chemical properties, the Hoechst affinity for DNA decreases with increasing osmolyte concentration. From the dependence of the equilibrium constant on the solution osmolality, 60 +/- 13 waters are acquired in the complex relative to the reactants. It is proposed that the osmotic stress technique is measuring weakly bound waters that are not measured via the heat capacity changes.

A molecular thermodynamic view of DNA–drug interactions: a case study of 25 minor-groove binders

Developing a molecular view of the thermodynamics of DNA recognition is essential to the design of ligands for regulating gene expression. In a first comprehensive attempt at sketching an atlas of DNA-drug energetics, we present here a detailed thermodynamic view of minor-groove recognition by small molecules via a computational study on 25 DNA-drug complexes. The studies are configured in the MMGBSA (Molecular Mechanics-Generalized Born-Solvent Accessibility) framework at the current state of the art and facilitate a structure-energy component correlation. Analyses were conducted on both energy minimized structures of DNA-drug complexes and molecular dynamics trajectories developed for the purpose of this study. While highlighting the favorable role of packing, shape complementarity, and van der Waals and hydrophobic interactions of the drugs in the minor groove in conformity with experiment, the studies reveal an interesting annihilation of favorable electrostatics by desolvation. Structural modifications attempted on the ligands point to the requisite physico-chemical factors for obtaining improved binding energies. Hydrogen bonds predicted to be important for specificity based on structural considerations do not always turn out to be significant to binding in post facto analyses of molecular dynamics trajectories, which treat thermal averaging, solvent, and counterion effects rigorously. The strength of the hydrogen bonds retained between the DNA and drug during the molecular dynamics simulations is approximately 1kcal/mol. Overall, the study reveals the compensatory nature of the diverse binding free energy components, possible threshold limits for some of these properties, and the availability of a computationally viable free energy methodology which could be of value in drug-design endeavors.